ACR/CS3 Black Point confusion

I'm having a problem where when I set the Black Point in ACR (aRGB conversion) to the point where it's starting to clip the histogram for the converted image in CS3 shows room before clipping (aRGB working space). Here are the Histograms that I see:

HOSTED PHOTO
please log in to view hosted photos in full size.

HOSTED PHOTO
please log in to view hosted photos in full size.

And here is the other histogram in CS3:

HOSTED PHOTO
please log in to view hosted photos in full size.

Is the above normal, or am I asking for too much from Adobe? Could I have some other setting wrong?

Any help or insight would be greatly appreciated

Strange, but true it seems.

Here's CS4:
Not a big difference, but a difference.

Also weird that I don't seem to get consistent clipping warnings in CS4.
If I drag the black point on the curve from 0 to the right, and back agian, it's almost as if there's a screen refresh problem

Again, not a big difference, but I'd expect the exact same image to give the exact same clipping warning

Man o man, I'm glad you replied René. This ones been driving me crazy because I edit in ACR by moving the black point so it barely clips, then open up in CS3 and I have to move the black point in curves some more (so it barely clips). Based on the image being a bit "flat" looking in ACR, and not after correcting it a bit more in CS3 ... I have a feeling ACR is the one that's off.

There's no way the CS3 color settings can cause this can it?

HOSTED PHOTO
please log in to view hosted photos in full size.

Nope.
I think ACR is a bit "conservative" if you adjust a slider while pressing Opt.
The "real" clipping warning is closer to what PS proper gives, but still a bit over-enthousiastic

I'll add this to your discussion regarding histograms.

The Histogram can be refreshed in PS, your cache states in preferences controls the refresh states and you can manually refresh. Do a google search for more info.

here's an interesting thread from another board (found via google)
https://www.naturescap​es.net …hp?f=2&t=144546​&p=1458448

I've been using the Alt-Slider approach for the clipping display for a few years ... it really is a great feature isn't it?

I was shooting RAW + jpeg with my D700 and wondering why Nikon's jpeg straight out of the camera (for a properly exposed image) got a black point that needed no adjustment in CS3, but when I processed the corresponding RAW it was off. And the result was really flat looking compared to the jpeg from the camera.

I've switched my approach to add +4 or +5 Black Point in ACR after I've moved the slider to the point where I see the tiniest bit of clipping (Alt-Slider). I then check the histogram and image shadow detail in CS3 just to make sure I didn't go too far.

I need to check, but I think there may be the same conservatism for highlights. Why Adobe did this is beyond me. Maybe it was to ensure that you can switch from aRGB to sRGB using "Convert to Profile" in Photoshop and not lose any data.

Palladium wrote in post #7259153
I'll add this to your discussion regarding histograms.

The Histogram can be refreshed in PS, your cache states in preferences controls the refresh states and you can manually refresh. Do a google search for more info.

here's an interesting thread from another board (found via google)
https://www.naturescap​es.net …hp?f=2&t=144546​&p=1458448

Thanks Palladium. Another good PS tip.

The issue here isn't a refresh problem though. It's an actual issue where the ACR histogram and clipping warning does not match the same histogram in PS. ACR tells you it's clipped where you'd have to go +4 further on the slider before you'd see any clipping when viewing the jpeg in CS3.

I don't recall it being this far off with other versions of ACR ... but maybe I just didn't notice it.

Good info direct from adobe about histograms and understanding the clipping warnings

http://livedocs.adobe.​com …5b-BBA1-627CF3D3ED26.html

Palladium wrote in post #7259339
Good info direct from adobe about histograms and understanding the clipping warnings

http://livedocs.adobe.​com …5b-BBA1-627CF3D3ED26.html

Thanks, but are you thinking René and I are looking at something wrong or can you see that what we are showing is actually an error/problem between Photoshop and ACR?

I tried to replicate this problem with a test image but couldn't do it. First I created this gradient so it is "safe" from clipping and blocking.

HTTP response: NOT FOUND | MIME changed to 'image/png'

HTTP response: NOT FOUND | MIME changed to 'image/png'

Peano wrote in post #7263595
I tried to replicate this problem with a test image but couldn't do it. First I created this gradient so it is "safe" from clipping and blocking.

HTTP response: NOT FOUND | MIME changed to 'image/png'

I saved that as a jpeg and opened first in ACR, then in PS, and compared the histograms. Both showed approximately the same gaps at either end of the scale -- just what you'd expect.

Then I opened it in ACR and pushed exposure and blacks until the first warnings of clipping and blocking showed:

HTTP response: NOT FOUND | MIME changed to 'image/png'

Checked both histograms again, and each showed slight clipping and blocking.

I'm not sure what can be concluded from this, since I started with a jpeg. Maybe starting with a raw file is what makes for the difference you're seeing.

EDIT:

I just tried with a raw file. Pushed exposure until the first clipping showed, and increase blacks until blocking just barely showed. Then opened in PS and checked the histogram. There were no obvious spikes at either end.

But when I opened a curves dialog box and checked "show clipping," both the clipping and the blocking showed. So at least in that respect, ACR and PS seem to be "seeing" the same thing, even though it isn't obvious when you view the PS histogram.

Start in RAW and push the black until clipping just shows and then open in PS and look at the histogram with curves and you'll see you have a bunch more room. It doesn't show more clipping in PS curves, it shows LESS.

I'm using aRGB for the RAW conversion and the PS colorspace, ACR 4.6 or higher for the RAW conversion. I don't recall seeing this issue with earlier versions of ACR, but I also used to process in sRGB ... what color space are you using?

René's example is spot on what I see.

Bob_A wrote in post #7264707
Start in RAW and push the black until clipping just shows and then open in PS and look at the histogram with curves and you'll see you have a bunch more room. It doesn't show more clipping in PS curves, it shows LESS.

I'm not seeing that in CS3. I've tried using sRGB and aRGB. In ACR 4.6 with a raw file, I increase blacks until the first traces of blocked shadows appear on the image. Then I open in PS, open curves, click "Show clipping, and click the left slider. The same traces of blocked shadows appear there. Sorry, but I'm just not seeing the difference you describe.